You are on page 1of 8

American Journal of Engineering Research (AJER)

2015

American Journal of Engineering Research (AJER)


e-ISSN : 2320-0847 p-ISSN : 2320-0936
Volume-04, Issue-01, pp-08-15
www.ajer.org
Research Paper

Open Access

DMP Packet Scheduling For Wireless Sensor Network


Pallavi Sawale1 , Dr.D.J Pete2
1

(Department of Electronics,Datta Meghe College of engineering/Mumbai University,India)


(Department of Electronics,Datta Meghe College of engineering/Mumbai University,India)

ABSTRACT : Most of the existing packet scheduling mechanisms of the wireless sensor network use First
Come First Served (FCFS) non pre emptive priority and pre emptive priority scheduling algorithms. The above
algorithms have high processing overhead and also long end-to-end data transmission delay. In FCFS concept
the data packet which is entering the node first will go out first from the node, and the packet which will enter
last will leave at last. But in FCFS scheduling of real time data packets coming to the node have to wait for a
long time period.In non pre emptive priority scheduling algorithm there is starvation of real time data packets
because once the processor enters the running state, it will not allow remove until it is completed, so there is
starvation of real time data packets. In pre emptive scheduling, starvation of non real time data packets, due to
continuous arrival of real time data. Therefore the data packets are to be schedule in multilevel queue. But the
multilevel queue scheduling scheme is not suitable for dynamic inputs, and hence the scheme is designed for
dynamically change in the inputs. The Dynamic Multilevel Priority (DMP) packet scheduling is the scheme for
dynamically changes in the inputs. In this scheme each node except the last level of the virtual hierarchy in the
zone based topology of wireless sensor network has three levels of priority queues. Real time data packets are
placed into highest priority queue and can preempt the data packets in the other queues. Non real time data
packets are placed into other two queues based on threshold of their estimated processing time. The leaf node
have two queues, one for real time data packet and another for non real time data packet since they do not
receive data from other nodes and thus reduces end to end delay. This scheme reduces the average waiting time
and end-to end delay of data packets.

KEYWORDS : Data waiting time, FCFS , non-preemptive priority scheduling,non-real-time,

packet

scheduling, pre-emptive priority scheduling,real-time, Wireless sensor network.

1. INTRODUCTION
Scheduling of different packets at the sensor nodes is very important as ensures the delivery of the data
packet on the priority basis. The sensed data may be real time or non real time. Highest priority should be given
to the real time data sense by the node compare to non real time data packet.[1] Sometimes the nodes may be put
to sleep mode, when there is no data packet available and as soon as the data packet arrives at the node is putted
into wake mode. This reduces the sensor node energy consumption. In FCFS scheduling scheme the data packets
are processed in order of their arrival time and therefore the data packet which is entering at the last will require a
long time to reach to a base station .However the emergency real time data should be reach to the base station
before the deadline expires. Since the emergency real time data packet should reach the base station with shortest
possible end-to-end delay. But most of the packets scheduling algorithms are neither dynamic nor suitable for
large scale applications as their scheduling is predetermine and they are static. Static means those cannot be
change as per the application requirement.[2] In the proposed dynamic multilevel priority packet scheduling
scheme hierarchical structure of node is used. The nodes that have same hop distance from the base station will
be considered to be located at the same hierarchical level. Processing of the data packets is done by TDMA
scheme. In TDMA scheme the time slot is assigned for each data packet. Three levels of priority queues are
defined. First priority will be given to real time data packet, second priority will be given to non real time data
packets that are sensed by remote node at lower level & third priority will be given to the non real time data
packet that are local sensed at the node.

www.ajer.org

Page 8

American Journal of Engineering Research (AJER)

2014

If the non real time data packets with same priority are arrived then they are processed using the shortest
job first scheduler scheme. The advantages of wireless sensor networks (WSNs) have lately become interesting
for industrial and factory automation, distributed control system, automotive systems and other kinds of network
embedded systems. The need for reduced cabling, faster setup times for equipment, reliable communication in
harsh areas and added mobility have triggered research on the use of wireless communication in industrial
systems. Even if WSNs provide a lot of benefits in the context of industrial communication, they also suffer
from a number of disadvantages. Wireless communication is characterized by its high error probability, leading to
the risk of causing severe problems for applications with strict reliability and timing requirements. Lost or
delayed data may cause industrial applications to malfunction. Transmission scheduling in the context of time
division multiple accesses (TDMA) can achieve robust and collision-free communication. Meanwhile, the
TDMA-based medium access control (MAC) protocols can provide quality of service (QoS) access to the
wireless network. Therefore a TDMA-based scheme is preferred. In TDMA MAC protocols, time is slotted into
intervals of equal length, which are called timeslots. The duration of one timeslot is equal to the time required to
transmit a packet and return an acknowledgement (ACK). A collection of timeslots that repeat cyclically are
grouped into super frames. The radio spectrum is divided into channels with small frequency bands, which can
realize parallel transmissions and enhance network throughput. Although these TDMA-scheduling approaches
can find minimum length schedule, minimum energy schedules or a fairness-based schedule, they do not account
for reliability. However reliability in the WSNs is particularly significant. Reliability expresses the probability of
successful packet delivery from a source node to the destination node. On each wireless link, a packet loss rate
(PER) of 1030% is common, which significantly decreases end-to-end reliability. An example shows that
assuming 10% PER and three transmission attempts on each link, 99% packets are received over one hop. After
ten hops, success probability is only 76 % [4] .An Energy aware Coverage based Node Scheduling scheme
(ECNS) is proposed to provide protection for sensors and guarantees network connectivity and desired coverage
level. ECNS enables each node to decide whether it is eligible to turn off to conserve energy through local
information exchange with its neighbors. Simulation results show that ECNS improves network performance with
respect to energy conservation, load balance and network lifetime. [6]

II. COMPARISON
Comparison of several existing WSN packet or task scheduling algorithms. The existing task scheduling
algorithm are based on several factors such as Deadline: - Packet scheduling schemes can be classified based on
the deadline of arrival of data packets to the base station (BS), which are as follows First Come First Served
(FCFS): Most existing WSN applications use First Come First Served (FCFS) schedulers that process data in the
order of their arrival times at the ready queue. In FCFS, data that arrive late at the intermediate nodes of the
network from the distant leaf nodes require a lot of time to be delivered to base station (BS) but data from nearby
neighboring nodes take less time to be processed at the intermediate nodes. In FCFS, many data packets arrive
late and thus, experience long waiting times.
Earliest Deadline First (EDF): -Whenever a number of data packets are available at the ready queue and each
packet has a deadline within which it should be sent to BS, the data packet. which has the earliest deadline is sent
first. This algorithm is considered to be efficient in terms of average packet waiting time and end-to-end delay.
Data, that have travelled the longest distance from the source node to BS and have the shortest deadline, are
prioritized. If the deadline of a particular task expires, the relevant data packets are dropped at an intermediate
node.
Packet Type: - Packet scheduling schemes can be classified based on the types of data packets, which are as
follows.
Real-time packet scheduling: Packets at sensor nodes should be scheduled based on their types and priorities.
Real-time data packets are considered as the highest priority packets among all data packets in the ready queue.
Hence, they are processed with the highest priority and delivered to the BS with a minimum possible end-to-end
delay.
Priority: Packet scheduling schemes can be classified based on the priority of data packets that are sensed at
different sensor nodes.

www.ajer.org

Page 9

American Journal of Engineering Research (AJER)

2014

Non-preemptive: In non-preemptive priority packet scheduling, when a packet t1 starts execution, task t1
carries on even if a higher priority packet t2 than the currently running packet t1 arrives at the ready queue. Thus
t2 has to wait in the ready queue until the execution of t1 is complete.
Preemptive: In preemptive priority packet scheduling, higher priority packets are processed first and can
preempt lower priority packets by saving the context of lower priority packets if they are already running.
Packet Type: Packet scheduling schemes can be classified based on the types of data packets, which are as
follows.
Real-time packet scheduling: Packets at sensor nodes should be scheduled based on their types and priorities.
Real-time data packets are considered as the highest priority packets among all data packets in the ready queue.
Hence, they are processed with the highest priority and delivered to the BS with a minimum possible end-to-end
delay.
Non-real-time packet scheduling: Non-real time packets have lower priority than real-time tasks. They are
hence delivered to BS either using first come first serve or shortest job first basis when no real-time packet
exists at the ready queue of a sensor node. These packets can be intuitively preempted by real-time packets.
Number of Queue: - Packet scheduling schemes can also be classified based on the number of levels in the
ready queue of a sensor node. These are as follows.
Single Queue: Each sensor node has a single ready queue.
All types of data packets enter the ready queue and are scheduled based on different criteria: type, priority, size,
etc.
Single queue scheduling has a high starvation rate.
Multi-level Queue: Each node has two or more queues. Data packets are placed into the different queues
according to their priorities and types. Thus, scheduling has two phases: (i) Allocating tasks among different
queues, (ii) scheduling packets in each queue. The number of queues at a node depends on the level of the node
in the network. For instance, a node at the lowest level or a leaf node has a minimum number of queues whilst a
node at the upper levels has more queues to reduce end-to-end data transmission delay and balance network
energy consumptions. Figure 1 illustrates the main concept behind multi-level queue scheduling algorithms

www.ajer.org

Page 10

American Journal of Engineering Research (AJER)


III.

2014

ASSUMPTIONS

The following assumptions are made to design and implement


DMP packet scheduling scheme.
Data traffic comprises only real-time and non-real-time data, e.g., real-time health data sensed by body sensors
and non-real-time temperature data.
All data packets (real-time and non-real-time) are of same size.
Sensors are time synchronized.
No data aggregation is performed at intermediate nodes for real-time data.
Nodes are considered located at different levels based on the number of hop counts from BS.
Timeslots are allocated to nodes at different levels using TDMA scheme.
The ready queue at each node has maximum three levels or sections for real-time data (pr1) non-real-time
remote data (pr2) and non-real-time local data (pr3). The length of data queues is variable. For instance, the
length of real-time data queue (pr1) is assumed to be smaller than that of non-real-time data queues (pr2 and
pr3). However, the length of the non-real-time pr2 and pr3 queues are same.
DMP scheduling scheme uses a multichannel MAC protocol to send multiple packets simultaneously.

IV.

TERMINOLOGIES

In this section, we define the following terminologies and factors that are used in designing the DMP
packet scheduling scheme.Routing Protocol: For the sake of energy efficiency and balance in energy
consumption among sensor nodes, we envision using a zone-based routing protocol [ 8,9]. In a zone based
routing protocol, each zone is identified by a zone head (ZH) and nodes follow a hierarchical structure, based on
the number of hops they are distant from the base station (BS). For instance, nodes in zones that are one hop and
two hops away from the BS are considered to be at level 1 and level 2, respectively. Each zone is also divided
into a number of small squares in such a way that if a sensor node exists in square S1, it covers all neighboring
squares. Thus, this protocol reduces the probability of having any sensing hole in the network even if the
neighboring squares of a node do not have any sensor node.TDMA Scheme: Task or packet scheduling at each
nodal level is performed using a TDMA scheme with variable-length timeslots. Data are transmitted from the
lowest level nodes to BS through the nodes of intermediate levels. Thus, nodes at the intermediate and upper
levels have more tasks and processing requirements compared to lower-level nodes. Considering this
observation, the length of timeslots at the upper-level nodes is set to a higher value compared with the timeslot
length of lower-level nodes. On the other hand, real-time and time critical emergency applications should stop
intermediate nodes from aggregating data since they should be delivered to end users with a minimum possible
delay. Hence, for real-time data, the duration of timeslots at different levels is almost equal and short.Fairness:
This factor ensures that tasks of different priorities get carried out with a minimum waiting time at the ready
queue based on the priority of tasks. For instance, if any lower priority task waits for a long period of time for
the continuous arrival of higher-priority tasks, fairness defines a constraint that allows the lower-priority tasks to
get processed after a certain waiting time.Priority: The priority of non-real-time data packets is assigned based
on the sensed location (i.e., remote or local) and the size of the data. The data packets that are received by node
x from the lower level nodes are given higher priority than the data packets sensed at the node x itself. However,
if it is observed that the lower priority non-real-time local data cannot be transmitted due to the continuous
arrival of higher priority non-real-time remote data, they are preempted to allow low-priority data packets to be
processed after a certain waiting period. Nevertheless, these tasks can be preempted by real-time emergency
tasks. In case of two same priority data packets the smaller sized data packets are given the higher priority.

V.

PROPOSED DMP PACKET SCHEDULING SCHEME

In non-preemptive packet scheduling schemes real-time data packets have to wait for completing the
transmissions of other non-real-time data packets. On the other hand, in preemptive priority scheduling, lowerpriority data packets can be placed into starvation for continuous arrival of higher priority data. In the multilevel
queue scheduling algorithm [], each node at the lowest level has a single task queue considering that it has only

www.ajer.org

Page 11

American Journal of Engineering Research (AJER)

2014

local data to process. However, local data can also be real-time or non-real time and should be thus processed
according to their priorities. Otherwise, emergency real-time data traffic may experience long queuing delays till
they could be processed. Thus, we propose a Dynamic Multilevel Priority (DMP) packet scheduling scheme that
ensures a tradeoff between priority and fairness.
Working Principle: Scheduling data packets among several queues of a sensor node is presented in Figure 2.
Data packets that are sensed at a node are scheduled among a number of levels in the ready queue. Then, a
number of data packets in each level of the ready queue are scheduled. For instance, Figure 2 demonstrates that
the data packet, Data1 is scheduled to be placed in the first level, Queue1. Then, Data1 and Data3 of Queue1
are scheduled to be transmitted based of different criteria. The general working principle of the proposed DMP
scheduling scheme is illustrated in Figure 3.

Fig. 2 Scheduling data among multiple queues.


The proposed scheduling scheme assumes that nodes are virtually organized following a hierarchical
structure. Nodes that are at the same hop distance from the base station (BS) are considered to be located at the
same level. Data packets of nodes at different levels are processed using the Time-Division Multiplexing Access
(TDMA) scheme. For instance, nodes that are located at the lowest level and the second lowest level can be
allocated timeslots 1 and 2, respectively. We consider three-level of queues, that is, the maximum number of
levels in the ready queue of a node is three: priority 1 (pr1), priority 2 (pr2), and priority 3 (pr3) queues. Realtime data packets go to pr1, the highest priority queue, and are processed using FCFS. Non-real-time data
packets that arrive from sensor nodes at lower levels go to pr2, the second highest priority queue. Finally, nonreal time data packets that are sensed at a local node go to pr3, the lowest priority queue. The possible reasons
for choosing maximum three queues are to process (i) real-time pr1 tasks with the highest priority to achieve the
overall goal of WSNs, (ii) non real-time pr2 tasks to achieve the minimum average task waiting time and also to
balance the end-to-end delay by giving higher priority to remote data packets, (iii) non-real-time pr3 tasks with
lower priority to achieve fairness by preempting pr2 tasks if pr3 tasks wait a number of consecutive timeslots
processed using FCFS. Non-real-time data packets that arrive from sensor nodes at lower levels go to pr2, the
second highest priority queue

.
Fig.3. Proposed dynamic multilevel priority scheduling scheme.

www.ajer.org

Page 12

American Journal of Engineering Research (AJER)

2014

Finally, non-real time data packets that are sensed at a local node go to pr3, the lowest priority queue.
The possible reasons for choosing maximum three queues are to process (i) real-time pr1 tasks with the highest
priority to achieve the overall goal of WSNs, (ii) non real-time pr2 tasks to achieve the minimum average task
waiting time and also to balance the end-to-end delay by giving higher priority to remote data packets, (iii) nonreal-time pr3 tasks with lower priority to achieve fairness by preempting pr2 tasks if pr3 tasks wait a number of
consecutive timeslots. In the proposed scheme, queue sizes differ based on the application requirements. Since
preemptive priority scheduling incurs overhead due to the context storage and switching in resource constraint
sensor networks, the size of the ready queue for preemptive priority schedulers is expected to be smaller than
that of the preempt able priority schedulers. The idea behind this is that the highest-priority real-time/emergency
tasks rarely occur. They are thus placed in the preemptive priority task queue (pr1 queue) and can preempt the
currently running tasks. Since these processes are small in number, the number of preemptions will be a few. On
the other hand, non real time packets that arrive from the sensor nodes at lower level are placed in the pre
emptable priority queue (pr2 queue). The processing of these data packets can be preempted by the highest
priority real-time tasks and also after a certain time period if tasks at the lower priority pr3 queue do not get
processed due to the continuous arrival of higher priority data packets. Real-time packets are usually processed
in FCFS fashion. Each packet has an ID, which consists of two parts, namely level ID and node ID. When two
equal priority packets arrive at the ready queue at the same time, the data packet which is generated at the lower
level will have higher priority. This phenomenon reduces the end-to-end delay of the lower level tasks to reach
the BS. For two tasks of the same level, the smaller task (i.e., in terms of data size) will have higher priority.
Moreover, it is expected that when a node x senses and receives data from lower-level nodes, it is able to
process and forward most data within its allocated timeslot; hence, the probability that the ready queue at a node
becomes full and drops packets is low. However, if any data remains in the ready queue of node x during its
allocated timeslot, that data will be transmitted in the next allocated timeslot.

VI.

PERFORMANCE EVALUATION

The simulation model is implemented using the java programming. It is used to evaluate the
performance of the proposed DMP packet scheduling scheme, comparing it against the FCFS, and Multilevel
Queue scheduling schemes. The comparison is made in terms of average packet waiting time, and end-to-end
data transmission delay. We use randomly connected Unit Disk Graphs (UDGs) on a surface of 100 meter 100
meter as a basis of our simulations. The number of simulated zones varies from 4 to 12 zones. Nodes are
distributed uniformly over the zones. The ready queue of each node can hold a maximum of 50 tasks. Each task
has a Type ID that identifies its type. For instance, type 0 is considered to be a real-time task. Data packets are
placed into the ready queue based on the processing time of the task. Moreover, each packet has a hop count
number that is assigned randomly, and the packet with the highest hop count number is placed into the highestpriority queue. We run the simulation both for a specific number of zones, and levels in the network until data
from a node in each zone or level reach BS. Simulation results are presented for both real-time data and all types
of data traffic. Table I presents simulation parameters, and their respective values.
TABLE I
Simulation Parameters And Their Values
Parameter
Network Size
Number of Nodes
Number of Zones
Base station position
Transmission Energy
Consumptions
Energy Consumption in
free space or air
Initial Node Energy
Transmission Speed
Propagation Speed

www.ajer.org

Value
100m X 100m
Maximum 200
4 - 12
55m X 101m
50 n Joule/bit
0.01 n Joule/bit/m2
2 Joule
250Kbps
198 106 meter/sec

Page 13

American Journal of Engineering Research (AJER)


VII.

2014

RESULT

End-End delay of real time data

Figure 4. End-to-end delay of real-time data over a number of zones


Figures 4. Demonstrates the end-to-end delay of all types of data traffic over a number of zones and levels.
From these results, we find that the DMP task scheduling scheme outperforms FCFS, and Multilevel Queue
scheduler in terms of end-to-end data transmission delay. This is because in the proposed scheme, the tasks that
arrive from the lower level nodes are given higher priority than the tasks at the current node. Thus, the average
data transmission delay is shortened. Furthermore, the average waiting time of a task contributes largely to the
experienced end-to-end data transmission delay. We are assigning task priority based on task deadline instead of
the shortest task processing time. To reduce processing over-head and save bandwidth, we could also consider
removing tasks with expired deadlines from the medium.

VIII. CONCLUSION AND FUTURE WORK


DMP task scheduler has better performance than the FCFS, and Multilevel Queue scheduler in terms of
average task waiting time, both for real-time tasks, and all types of tasks. Using the concept of three-level
priority queues at each node, the proposed DMP task scheduling scheme allows different types of data packets
to be processed based on their priorities. Since real time and emergency data should be processed with the
minimum end-to-end delay, they are processed with the highest priority, and can preempt tasks with lower
priorities located in the two other queues. On the other hand, in existing multilevel queue schedulers, a task with
the highest hop count is given the highest priority. Hence, real-time tasks are prioritized over other task types
only if their hop counts are higher than those of non-real-time tasks. Moreover, in FCFS and multilevel queue
schedulers, the estimated processing time of a task is not considered when deciding the priority of a task. Thus,
FCFS and Multilevel Queue schedulers exhibit longer task waiting times and end-to-end delays, in comparison
to the DMP task scheduling scheme. If a real-time task holds the resources for a longer period of time, other
tasks need to wait for an undefined period time, causing the occurrence of a deadlock. This deadlock situation
degrades the performance of task scheduling schemes in terms of end-to-end delay. Hence, we would deal with
the circular wait and preemptive conditions to prevent deadlock from occurring.

REFERENCES
[1]
[2]
[3]
[4]

Nidal Nasser, Lutful Karim, and Tarik Taleb Dynamic Multilevel Priority Packet Scheduling Scheme for Wireless Sensor
Network, IEEE Transactions On Wireless Communications, VOL. 12, NO. 4, APRIL 2013, pp. 1448-1459.
Ali Sharifkhani, and Norman C. Beaulieu, A Mobile-Sink-Based Packet Transmission Scheduling Algorithm for Dense Wireless
Sensor Networks, IEEE Transactions On Vehicular Technology, VOL. 58, NO. 5, JUNE 2009, pp. 2509-2518.
Nilesh Khude, Anurag Kumar, Aditya Karnik, Time and Energy Complexity of Distributed Computation of a Class of Functions in
Wireless Sensor Networks, IEEE Transactions On Mobile Computing, VOL. 7, NO. 5, MAY 2008, pp. 617-631.
Xiaoling Zhang, Wei Liang, Haibin Yu, Xisheng Feng, Reliable transmission scheduling for multi-channel wireless sensor

www.ajer.org

Page 14

American Journal of Engineering Research (AJER)


[5]
[6]
[7]
[8]
[9]
[10]

2014

networks with low-cost channel estimation, IET Communications, ISSN1, 2013, Vol. 7, pp. 7181.
F. Liu, C. Tsui, and Y. J. Zhang, Joint routing and sleep scheduling for lifetime maximization of wireless sensor networks, IEEE
Trans. Wireless Communication, vol. 9, no. 7, pp. 22582267, July 2010.
J.Liu, N.Guo, and S. He, An energy-aware coverage based node scheduling scheme for wireless sensor networks, in Proc. 2008
International Conf. Young Comput. Scientists, pp. 462468.
S. Paul, S. Nandi, and I. Singh, A dynamic balanced-energy sleep scheduling scheme in heterogeneous wireless sensor network,
in Proc. 2008 IEEE International Conf. Netw., pp. 16, 2008.
O. Khader, A. Willig, and A. Wolisz, Distributed wakeup scheduling scheme for supporting
periodic traffic in wsns, in Proc.
2009 European Wireless Conf., pp. 287292.
B. Nazir and H. Hasbullah, Dynamic sleep scheduling for minimizing delay in wireless sensor network, in Proc. 2011 Saudi
International Electron., Communications Photon. Conf., pp. 15.
D. Shuman and M. Liu, Optimal sleep scheduling for a wireless sensor network node, in Proc. 2006 Asilomar Conf. Signals,
Syst. Comput., pp 13371341.

www.ajer.org

Page 15

You might also like